Compressor Capacity Selection & System Sizing

REDDY COMPRESSORS

Future revisions must be versioned.

The right capacity is not the biggest machine you can justify

Compressor sizing is the controlled conversion of a system requirement into an equipment duty.

The answer should not be created by adding a familiar percentage to an uncertain CFM figure or automatically inheriting the existing compressor size. A defensible selection preserves demand evidence, pressure, operating pattern, site conditions, treatment boundary, control philosophy and availability strategy.

The objective is sufficient, controllable and maintainable capacity—not maximum installed nameplate capacity.

1 — Freeze the Design Basis

If inputs remain unresolved, equipment comparison remains conditional.

2 — Size Defined Operating Cases

Keep normal sustained demand, normal peaks, low demand, critical-process cases and future scenarios separate. Neither the highest observed point nor the average automatically defines continuous capacity.

3 — Pair Capacity With Pressure

A candidate must support required flow at required pressure. High maximum pressure alone does not prove flow at that pressure.

4 — Keep Airflow Bases Reconciled

Calculate capacity surplus or shortfall only after demand and supply flow definitions/reference conditions are compatible.

5 — Separate Sustained Capacity From Transients

Short peaks may be supported by storage, distribution and control response. Peak magnitude, duration, allowable pressure change and recharge requirement must be established.

6 — Separate Productive From Avoidable Demand

Leakage, artificial demand and inappropriate uses should not automatically become permanent design capacity. Show current and post-improvement scenarios when savings are not yet verified.

7 — Existing Compressor Size Is Evidence

The installed machine can inform the study but does not define current demand. Plant requirements and machine condition may have changed.

8 — Connected Load Is Not Compressor Capacity

Sum connected consumers only when simultaneity supports that design case. Do not use an unsupported diversity assumption either.

9 — Design Provision Needs a Reason

Capacity provision may address measurement uncertainty, committed growth, maintenance strategy or another named need. State each reason separately.

10 — Avoid Compounded Hidden Margins

Audit conservatism already present in consumer estimates, branch totals and procurement requirements before adding further reserve.

11 — Future Growth Needs Scenarios

Committed growth belongs in a defined future case; probable and possible growth remain scenarios. Infrastructure can sometimes be prepared without buying all future capacity today.

12 — Redundancy Is Not Oversizing

Reliability strategy must define what service survives an outage. Standby, N+1, shared reserve and rental contingency are different strategies.

13 — Maintenance State Matters

Assess the plant condition while a compressor is unavailable for planned maintenance and whether reduced production or alternative supply is acceptable.

14 — Treatment Can Limit Effective Capacity

Dryers, filters and treatment equipment have flow capability and pressure loss. Compressor sizing must reconcile the treatment train.

16 — Site Conditions Matter

Elevation, ambient temperature, humidity, ventilation and cooling can affect performance or limits. Use manufacturer-backed information for the actual configuration.

17 — Duty Pattern Matters

Continuous, intermittent, cyclic and highly variable demand can favour different machine/control arrangements. A headline capacity does not establish unrestricted duty.

18 — Fixed-Speed and VSD Solve Different Problems

Neither is automatically superior. Compare actual load distribution, supported turndown, pressure requirement, controls and performance.

19 — Minimum Capacity Matters

Wide demand variation can create poor control if machines cannot reduce output sufficiently. Assess minimum supported capacity, unload behaviour, storage and sequencing.

20 — Multiple Compressors Form a Capacity Portfolio

Assess combinations of available machines across demand states and identify base, trim, standby and maintenance roles. Installed totals alone are insufficient.

21 — Sequencing Changes Usable Capacity

Poor sequencing can leave machines partly loaded or unloaded while pressure remains unstable. Develop capacity and control architecture together.

22 — Motor Power Is Not Capacity

Motor kW or hp does not establish delivered airflow. Compare supported delivered capacity at the required pressure.

23 — Specific Energy Is Not a Sizing Rule

Energy performance matters only after required air service is established. Compare energy across representative operating states, not one favourable point.

24 — Avoid One Catalogue Maximum

Identify exact model, pressure version, motor/drive, cooling arrangement, controls and options. Product-family maximums can describe different variants.

25 — Temporary Capacity Is a Separate Case

Shutdown, commissioning, seasonal or breakdown requirements can justify temporary air rather than permanent installed capacity.

26 — Rental Contingency Is Not Installed Redundancy

Availability, transport, connection, treatment and response arrangements distinguish rental contingency from installed standby.

27 — Capacity Acceptance Needs a Boundary

Define whether acceptance is at package outlet, downstream of treatment, plant header or another point and use a compatible measurement method.

28 — Candidate Comparison Needs More Than CFM

Compare configuration, flow at pressure, range, site applicability, treatment interface, controls, drive requirements, cooling, maintenance, evidence and availability.

29 — Availability Must Not Rewrite the Requirement

An available substitute still has to satisfy the frozen duty. Technical suitability and commercial availability remain separate.

30 — Changes Reopen Selection

Material changes in production, pressure, quality, hours, distribution, treatment, site conditions or future-load commitment require controlled reselection.

31 — Capacity Design-Case Register

32 — Candidate Compressor Comparison Matrix

33 — Technical Verification Gate

35 — Frequently Asked Questions

How many CFM does my factory need?

Establish demand profile, simultaneity, peaks, pressure and airflow basis first. Capacity follows the verified design cases.

Should I add 20% or 30% extra capacity?

There is no universal margin. Add provisions only for stated engineering reasons.

Should the highest measured peak determine size?

Not automatically. Review duration, allowable pressure variation, storage, distribution and control response.

Can a receiver reduce compressor size?

Storage can support defined transients but does not replace sustained source capacity.

Is a bigger compressor safer?

Not necessarily. Oversizing can create inefficient control, cycling or poor turndown.

Should I copy the existing compressor size?

Use it as evidence, not as the sizing rule. Reassess current demand and pressure.

Is VSD always best for changing demand?

No. Compare actual load distribution, turndown, pressure control and alternative arrangements.

How should future expansion be included?

Separate committed, probable and possible loads and use scenarios.

What does redundancy mean?

The required service remains available under the defined outage case; the strategy depends on criticality and machine portfolio.

Can I choose rental air by matching CFM?

No. Reconcile flow basis, pressure, treatment, connection, site conditions and operating pattern.

What should I send for sizing?

Demand profile, pressure, airflow basis, production schedule, existing machines, treatment/distribution, site conditions, future loads and continuity requirements.

38 — Final CTA

DO NOT BUY THE BIGGEST COMPRESSOR. DEFINE THE DUTY, THE PRESSURE, THE OPERATING CASES AND THE CONTINUITY REQUIREMENT—THEN SELECT CAPACITY.

Secondary CTAs: Submit Plant Demand Data • Review Working Pressure • Compare Compressor Capacity • Plan Expansion / Redundancy

Checkpoint Record

Future revisions require a new version identity.

V2 expands capacity selection, sizing, transient duty, redundancy, multi-compressor architecture, lifecycle evaluation, worked examples and publishing controls without deleting V1.

V2 STATUS: STAGING READY — TECHNICAL / EDITORIAL VERIFICATION GATED

V2.1 — Capacity Selection Starts After Demand and Pressure Are Controlled

The real question is what supported delivered capacity is required at each controlled pressure case, for how long, with what air quality, continuity requirement and future obligation. If upstream inputs remain unresolved, equipment comparison remains conditional rather than hiding uncertainty inside a larger machine.

V2.2 — Convert Demand Cases Into Source-Duty Cases

Carry forward low-demand, normal sustained, normal peak, critical-service, maintenance/outage and committed-growth cases. Each retains flow basis, pressure, quality, duration, production context and evidence status. The compressor plant must satisfy the portfolio of cases, not one headline maximum.

V2.3 — Capacity at Pressure Is the Fundamental Pair

Every candidate capacity must be paired with pressure. Never combine maximum flow from one operating point with maximum pressure from another. Record exact model/configuration, supported delivered flow, pressure, site condition and evidence source.

V2.4 — Sustained Versus Transient Duty

A 20-second peak and an eight-hour load should not automatically create the same capacity decision. Short peaks require analysis of net flow deficit, useful storage, allowable pressure decay, source response and recharge. Once an event is sustained beyond useful storage, source capacity must ultimately carry it.

V2.5 — Worked Sustained-Capacity Example

Candidate A supports 24.0 m3/min at that pressure and site condition. Nominal surplus is 2.0 m3/min before treatment consumption, uncertainty or future provision. This is not automatically redundancy or a recommended margin.

V2.6 — Worked Transient-Peak Logic

Illustrative only: sustained demand is 22 m3/min and a cyclic event reaches 30 m3/min for 15 seconds. If source capacity during the event is 24 m3/min, nominal deficit is 6 m3/min for 15 seconds, equivalent to 1.5 m3 reference-volume air before storage-state reconciliation. Receiver sizing still requires starting/minimum absolute pressure, system volume, temperature assumptions and recharge strategy.

V2.7 — Avoidable Demand Is Not Permanent Productive Demand

Leakage, artificial demand, inappropriate blowing and restrictions can inflate current demand. Maintain CURRENT AS-IS and VERIFIED POST-IMPROVEMENT cases where appropriate. Do not purchase against unverified savings, but do not silently institutionalise known waste either.

V2.8 — Decompose Design Margin

Replace a generic safety factor with named provisions for measurement uncertainty, committed growth, process variability, deterioration or another identified need. Maintain a margin ledger so conservatism at consumer, branch, plant and procurement levels does not compound invisibly.

V2.9 — Sensitivity Analysis Beats Arbitrary Oversizing

If sustained demand could reasonably be 20, 22 or 24 m3/min, test candidates across that range instead of selecting 30 m3/min because uncertainty feels uncomfortable. Sensitivity analysis reveals which missing measurement actually changes the equipment decision.

V2.10 — Future Growth as Scenarios

Separate committed, probable and possible growth. Committed loads can form a future design case; probable/possible loads can influence headers, electrical infrastructure, controls, spare bays and space without forcing all speculative capacity into today's machine.

V2.11 — Reserve Is Not Redundancy

Reserve is capacity above a design case with intended equipment available. Redundancy means defined service survives a specified outage. Prove redundancy by removing the unavailable machine and comparing remaining supported capacity, pressure, quality and controls with the critical-service case.

V2.12 — N+1 Must Define N

N+1 is incomplete unless N corresponds to a defined duty. The remaining combination after one unit is unavailable must still satisfy the agreed flow, pressure, quality and control requirements. Do not use N+1 as a label without the outage case.

V2.13 — Maintenance Is an Operating Case

Define expected production while a compressor is under planned maintenance. Some plants require full output; others accept controlled reduction. Include remaining source capacity, treatment capacity, controls, temporary-air options and process restrictions.

V2.14 — Low Demand Can Disqualify Oversizing

A machine selected for a future maximum may spend years at today's lower load. If minimum supported capacity exceeds low demand, cycling, unloading or poor turndown can result. The selected architecture must be capable at the top and controllable at the bottom.

V2.15 — VSD Turndown Needs Exact Evidence

Obtain supported minimum/maximum flow at relevant pressure and representative performance across the range. Do not assume linear airflow, constant efficiency or operation down to zero. Below the supported range, the package may unload, stop or require another control strategy.

V2.16 — Fixed-Speed Machines Can Form Effective Stages

Multiple fixed-speed compressors can create discrete capacity steps. With suitable storage and sequencing, this can suit some load profiles. Evaluate loaded performance, unloaded power, switching frequency, pressure band and combinations rather than treating fixed-speed as inherently obsolete.

V2.17 — Base-and-Trim Architecture

Base machines carry sustained load; a trim machine follows variability. The trim machine may be VSD or another suitable architecture. Test the split against low, normal and peak demand so the trim unit is neither chronically below its useful range nor too small to stabilise the system.

V2.18 — Multi-Compressor Combination Matrix

For each demand band, list expected running machines, supported capacity at pressure, control role and loading state. Repeat for maintenance and failure cases. Total installed nameplate capacity does not reveal whether useful intermediate combinations exist.

V2.19 — Treatment Can Be the Bottleneck

Dryers, filters and separators must support relevant flow and inlet conditions. Adsorption-dryer purge can increase source demand. Reconcile compressor output with net useful treated-air delivery at the selected boundary.

V2.20 — Distribution Can Be the Bottleneck

More compressor capacity cannot correct a restrictive main, branch, regulator or hose. If source pressure is stable while remote pressure collapses, distribution engineering takes priority over additional source capacity.

V2.21 — Site Conditions Change Performance

Elevation, ambient temperature, room temperature, cooling-water conditions and ventilation can affect performance or limits. Use manufacturer-backed data for the exact model. Do not apply a universal altitude or temperature derating.

V2.22 — Cooling and Ventilation Enable Capacity

A compressor meeting duty on paper can trip or degrade in a hot, recirculating room. Carry heat rejection, cooling airflow/water and installation limits into compressor-room engineering.

V2.23 — Electrical Infrastructure Must Support the Portfolio

Additional/larger machines can require switchgear, cabling, transformer capacity, starting/drive provisions, protection and harmonic/EMC review where relevant. Technical compressor fit is not installation readiness.

V2.24 — Energy Comparison Uses Representative Load Points

Full-load specific energy does not predict annual energy for a variable plant. Evaluate loaded power, unloaded behaviour, VSD part-load performance, purge/treatment consumption and sequence across representative demand states.

V2.25 — Capacity Acceptance Boundary

State whether capacity is accepted at package outlet, downstream of treatment, plant header or another point. Test method, reference basis and pressure must match the claim. Do not compare downstream plant flow blindly with upstream package FAD.

V2.26 — Field Acceptance Versus Certified Performance

A field test can verify system behaviour without reproducing formal factory test conditions. Document instruments, stabilisation, pressure, temperature, site conditions and conversion method, and describe accurately what the result proves.

V2.27 — Normalise Candidate Scope

Compare exact configuration, supported flow-pressure points, control range, cooling, drive, treatment interface, controls, maintenance, commissioning, warranty/service scope and evidence. A compressor-only quotation is not directly comparable with a complete package.

V2.28 — Keep Availability Separate From Technical Fit

Urgent stock availability matters commercially but should not rewrite the duty. Maintain separate TECHNICAL FIT and AVAILABILITY fields. Re-run the comparison for any substitute.

V2.29 — Rental Capacity as Designed Contingency

Rental can reduce installed standby capital for some plants but adds mobilisation, connection, treatment, fuel/electrical, site and availability dependencies. Pre-engineer temporary connections and define the rental duty rather than calling rental possibility installed redundancy.

V2.30 — Capacity Selection Failure Modes

Common errors include sizing from motor kW, summing all connected loads, using average demand only, sizing to one maximum logger point, compounding safety factors, ignoring low-load behaviour, treating nameplate total as usable capacity, ignoring treatment purge and selecting from stock availability.

V2.31 — Lifecycle Re-Selection Triggers

Reopen sizing when production, pressure, treatment, distribution, site conditions, operating hours or committed expansion changes materially. Preserve the prior basis and issue a controlled revision.

V2.33 — Premium Capacity Design-Case Register

V2.34 — Multi-Compressor Portfolio Matrix

V2.35 — Capacity Selection Verification Gate

Demand cases are revision-controlled and representative.

Demand and candidate supply use reconciled airflow bases.

Every capacity value is paired with required pressure.

Sustained and transient duties are separated.

Avoidable demand is not silently institutionalised.

Every capacity provision has a named reason.

Margins are checked for compounding.

Low-load behaviour and turndown are assessed.

Treatment and distribution constraints are visible.

Site/cooling conditions are verified for exact candidates.

Multi-compressor combinations cover normal and outage cases.

Redundancy is proven against a defined critical-service case.

Acceptance boundary and evidence method are defined.

Technical fit and commercial availability remain separate.